Objectives: To develop valid and reliable hospital discharge criteria and a scoring system that would be used to assess when a patient should be discharged from perioperative physical therapy (PT) care.
Design: We developed the postoperative physiotherapy discharge scoring tool (POP-DST), a tool composed of objective criteria and a scoring system that would be used to determine when a patient should be discharged from perioperative PT. It is a composite score of the following five subcategories: mobility; breath sounds; secretion clearance; oxygen saturation; and respiratory rate. The score for the POP-DST ranges from 6 to 15, with a score of > 13 indicating readiness for discharge. We examined the content validity of the the POP-DST using focus groups and a mailed survey. To determine interrater reliability, two therapists, who were blinded to each other's scores, assessed postsurgical patients. Validity was examined by comparing the decision to discharge based on the score on the POP-DST to the decision to discharge according to the therapist's judgment. In addition, subjects who were discharged from PT were followed-up 7 to 10 days later to determine whether they had developed any subsequent respiratory problems.
Patients: One hundred four surgical patients were assessed to determine the reliability and validity of the POP-DST. For the ability of the test to detect postoperative complications following discharge from PT, 204 surgical patients were followed-up after discharge from PT.
Results: Interrater reliability was moderately high (intraclass correlation coefficient = 0.76; r = 0.77). There was strong agreement between the decision to discharge the patient from PT based on the tool criteria compared to the therapist's judgment (kappa range, 0.91 to 0.96). The ability of the POP-DST to predict those patients who would not develop complications postoperatively was 94%.
Conclusion: The results indicate that the POP-DST would facilitate clinical decision making related to PT discharge planning in postsurgical populations. The instrument demonstrated strong content validity and predictive validity, as well as high levels of interobserver agreement. This tool should be considered as a work in progress until it is more fully validated.